GLACIATION OF LOWER COPPER RIVER 467 East of the driftless area the conditions are likewise unmistakable. The oversteep-ened western wall of the former Copper River fiord between Fickett Glacier and Flag Point shows that the trunk glacier of the Copper River valley extended south of Flag Point. Saddlebag Glacier valley also shows unmistakable evidence of former expansion as does the valley of Salmon Creek and McKinley Lake. Between these regions is one where absence of striae on the rock ledges and of erratics in the soil seems to point to absence of invasion by glacial ice. The incoherent material forming the flat floor of the region is all fine silt. The new railway cuts afford ample opportunity to find foreign bowlders, but none were found. The western boundary of the driftless area is well-defined. On the railway it is almost eight miles west of the Copper River, where a well marked terminal moraine of Sheridan Glacier crosses the track, 625 feet east of Mile Post 19. The eastern boundary is le,ss certain. No signs of glaciation are visible along the railway in the eight miles eastward to Flag Point, and it is uncertain how far outside the mountain front the Saddlebag and McKinley Lake Glaciers expanded. The cause of this driftless area was that the expanded Scott and Sheridan Glaciers, which apparently united to form a piedmont ice tongue, probably extended southward far enough to coalesce with the trunk glacier from the Copper River canyon. In the angle between these glaciers and the mountain front there is a small driftless area, not overridden by the maximum glacial advance. Vegetation and the Former Advance of Glaciers in the Copper River Canyon. Besides the interesting relationships of trees and shrubs growing upon the dirt-covered ice of Miles, Grinnell, Allen, Heney, and several other glaciers, as already described, the vegetation in the Copper River region tells a specific story of its recent glacial history. The Copper River delta and the southern part of the canyon support a dense coniferous forest, except where treelessness is caused (a) by ocean water on the tide flats and salt marshes, or (b) by aggrading glacial streams on the outwash gravel plains. The spruces are abundant from the ocean up to McPherson Glacier, but not north of there, though there is a thick forest of cottonwoods and alders, and one baby spruce was seen near Miles Glacier bridge. From this point northward no conifers were observed in the Copper River canyon south of the Tasnuna and Bremner Rivers. North of Tasnuna River, however, there are spruces along the canyon bottom and its lower slopes, but in the Copper River basin north of Wood Canyon and near Chitina there is thick, mature, spruce forest. The distribution of vegetation, therefore, shows a practical absence of spruces and hemlocks for nearly 40 miles along the Copper River, while at either end there is dense, mature, coniferous forest. The 40 mile strip which has no conifers is not barren, but is thickly covered with cottonwood and alder, wherever the slopes, the glaciers, and the glacial streams permit. Throughout the region the vegetation is terminated upward by a normal timberline. To account for this distribution we postulate a recent period of expansion of the glaciers between Tasnuna River and McPherson Glacier. The retreat from this expanded stage was long enough ago for occupation of the 40 mile strip by alders and cottonwoods and for a commencement of occupation by the conifers, as in the case of the young spruce near Miles Glacier bridge.